This research seeks to extend transactional theories of human development into the realms of complex family systems and adolescent health risk behaviors, specifically substance use and risky sexual activity, two sets of behaviors that carry substantial psychosocial and physical health risks and yet are increasingly common. Analytically, the research seeks to assess (1) how parenting practices and family processes serve as protective factors to inhibit youth engagement in health risk behaviors, and (2) how youth risk behaviors alter family processes and parenting behaviors across a range of family structures. The proposed research has five primary goals. First, this research will extend traditional transactional models into more normative adolescent risk behaviors, including substance use and risky sexual behaviors. Second, analyses will extend the focus on parenting to formerly unexplored arenas of family processes, namely the regularity of family routines, in addition to assessing parents' warmth and hostility and parental monitoring. Third and fourth, analyses will expand the conceptualization of parents to more fully explore complex family systems. Analytic models will separately assess parenting behaviors of mothers and fathers, and of resident versus nonresident and biological versus social parents. Fifth, analyses will explore the role of siblings as they influence both parents and other adolescents in the family. This research will draw data from two large, longitudinal survey studies of adolescents and families, the National Longitudinal Survey of Youth 1997 cohort (NLSY97), a nationally representative sample, and the Three-City Study, a representative sample of low-income, primarily minority families. Both studies follow youth from early through late adolescence. In order to maximally exploit the strengths of the data, three primary analysis techniques will be employed: multilevel growth models with time-varying predictors, paired multilevel models which control for non-independence of family members, and sibling fixed effects regression models. In sum, the proposed research seeks to expand substantially understanding of how family processes and adolescent engagement in risk behaviors co-occur by assessing reciprocal relationships between these two factors over the course of adolescence. Adolescent substance use and risky sexual activity have become increasingly normative, yet carry substantial potential physical and psychosocial costs. This research seeks to expand our understanding of the protective family factors which discourage adolescent risk behaviors, and of how risk behaviors alter the consistency and effectiveness of parenting and family processes. Following youth from early through late adolescence, results will inform understanding of the development of health risk behaviors through the central transition periods leading up to adulthood. [unreadable] [unreadable] [unreadable]